PROJECT SUMMARY The brain is by far the most complex organ, with hundreds or even thousands of distinct cell types intermingled in a tissue that appears, at first glance, to be fairly homogeneous. This presents a problem for biochemical studies of specific cell types, since the protein content of a given cell type is difficult or impossible to purify from the surrounding tissue. For example, in the striatum, medium spiny neurons of the direct vs. indirect pathway (dMSNs vs. iMSNs) express different molecular markers, and cell-type-specific activation or silencing produces opposing behavioral outputs. In the context of drug addiction, silencing iMSNs enhances the motivation to self- administer drugs, while silencing dMSNs reduces drug-seeking behavior. Treatment approaches that could correct the extensive molecular alterations caused by drugs of abuse are promising therapeutic strategies. However, given the opposing behavioral outputs of iMSNs vs. dMSNs, a drug that reduces drug-seeking behavior via one pathway could enhance the same behavior via the other pathway. Identification of cell-type specific targets would increase the possibility of modulating specific striatal pathways and behaviors, but our inability to purify iMSN vs. dMSN protein precludes such exploratory experiments. Here, we propose to develop a sorting approach based on a common Immunology technique, Magnetic Cell Sorting (MACS). We will express, in the dendritic spines of iMSN or dMSN neurons, an inert, extracellular ?TAG? that is recognized by highly specific, commercially available MACS sorting reagents. Cell-type-specificity will be achieved using a viral dual-injection strategy, in which a DIO/FLEX vector containing an inverted TAG construct will be injected into the striatum, and a retrograde Cre virus will be injected into the Ventral Tegmental Area (VTA) or the Ventral Pallidum (VP) to label neurons of the direct or indirect pathway, respectively. Sorting will be accomplished using metal-conjugated antibodies and a magnetized column, allowing rapid isolation of a large amount of biomaterial. In aim 1, we develop and validate the TAG system in the rat. In Aim 2, we label iMSNs and dMSNs in rats, and use an extended intermittent access herion self-administration paradigm to produce groups of rats that express low- and high-levels of addict-like behavior (as well as food reward for control). We will then isolate iMSNs and dMSNs from rats expressing high and low addict-like behavior, and use a quantitative mass spectrometry approach to identify proteins that are differentially expressed in iMSNs or dMSNs in rats showing high- vs. low-levels of addict-like behavior. Follow-up work beyond the scope of the grant will explore the potential of modulating the expression of these proteins in rats expressing addict-like behavior, with the goal of reducing drug-seeking behavior by specifically targeting cells of the iMSN or dMSN pathway. More generally, our tagging-and-sorting approach could be widely applicable to the study of cell-type- specific protein expression in the brain.